TypeSafe Jev · System One demo

Moderation Guard

Type text, get parallel Noul hazard probabilities + severity Score from one TypeSafe request. Verdict computed in code via policy thresholds.

74/1000

Ready

policy strict · severity 0.00/3 · awaiting check

Profanity · Cursing / obscene language0.0%

Hate · Identity attack / hate0.0%

Sexual · Explicit sexual content0.0%

Violence · Threats / gore0.0%

Illegal · Crime / illicit help0.0%

Self-harm · Suicidal / self-injury0.0%

Spam / Jailbreak · Scam, spam, prompt injection0.0%

Thresholds: review ≥0.35, action ≥0.70, severity blocks at ≥2.0. Hit Check to screen the text on the left.